Probability based health claims processing

Abstract

Systems and methods herein describe probability-based health claims processing. The described systems and methods access a plurality of pharmacy claims, determine an aggregate rating of the pharmacy claims based on pharmacy claims data, submit a first subset of pharmacy claims to a first pharmacy claims approval system, submit a second subset of pharmacy claims of the plurality of pharmacy claims to a second pharmacy claims approval system, receive, from the first pharmacy claims approval system, a first set of decisions for the first subset of pharmacy claims, and, receive from the second pharmacy claims approval system, a second set of decisions for the second subset of pharmacy claims.

What is claimed is: 
     
       1. A method comprising:
 accessing a plurality of pharmacy claims, wherein each pharmacy claim in the plurality of pharmacy claims includes a set of pharmacy claim data; 
 for each pharmacy claim in the plurality of pharmacy claims, determining a respective aggregate rating of the pharmacy claim based on the set of pharmacy claim data; 
 based on the respective aggregate rating of the plurality of pharmacy claims, selectively processing each of the plurality of pharmacy claims, wherein:
 the selectively processing is performed using at least one of:
 a first pharmacy claims approval system including a real-time claim analysis, or 
 a second pharmacy claims approval system including a statistical-model-based claim analysis, and 
 
 the selectively processing includes:
 determining whether a first rating associated with a first subset of pharmacy claims has met a ratings threshold; 
 in response to determining that the first rating has not met the ratings threshold, submitting the first subset of pharmacy claims to the first pharmacy claims approval system; 
 determining whether a second rating associated with a second subset of pharmacy claims has met the ratings threshold; and 
 in response to determining that the second rating associated with the second subset of pharmacy claims has met the ratings threshold, submitting the second subset of pharmacy claims to the second pharmacy claims approval system including a neural network; 
 
 
 submitting a portion of the second subset of pharmacy claims to the first pharmacy claims approval system; 
 receiving, from the first pharmacy claims approval system:
 a first set of decisions for the first subset of pharmacy claims, and 
 a set of validation decisions for the portion of the second subset of pharmacy claims; 
 
 storing training data including:
 the portion of the second subset of pharmacy claims, and 
 the set of validation decisions; 
 
 training the neural network of the second pharmacy claims approval system based on the training data to generate claims decisions for each of the plurality of pharmacy claims in the training data, wherein:
 the second pharmacy claims approval system is trained to generate a second set of decisions for the second subset of pharmacy claims, and 
 the neural network is trained by performing training operations including:
 accessing the training data; 
 extracting features from the training data to generate predictions by finding correlations among the extracted features of the plurality of pharmacy claims that affect the set of validation decisions, wherein the predictions include a probability that a given pharmacy claim of the plurality of pharmacy claims will be approved; and 
 building the neural network of the second pharmacy claims approval system to generate new predictions for new pharmacy claims in response to analyzing the training data; 
 
 
 receiving, from the second pharmacy claims approval system, a second set of decisions for the second subset of pharmacy claims; 
 validating the second set of decisions based on the set of validation decisions; 
 causing display of the first set of decisions and the second set of decisions on a graphical user interface; 
 in response to determining that a respective claim of the plurality of pharmacy claims has been approved:
 generating a display corresponding to the respective claim with a first visual indicator, 
 identifying a prescription drug associated with the respective claim, and 
 fulfilling the prescription drug without operator intervention by:
 automatically controlling movement of a set of containers relative to an automated dispensing device, and 
 automatically dispensing, via the automated dispensing device, the prescription drug into a container of the set of containers; and 
 
 
 in response to determining that the respective claim has not been approved:
 generating a display corresponding to the respective claim with a second visual indicator, and 
 displaying a notification. 
 
 
     
     
       2. The method of  claim 1 , wherein the set of pharmacy claim data includes at least one of drug data associated with a drug, physician data associated with a physician, pharmacy data associated with a pharmacy, benefit group data associated with a benefit group, or member data associated with a member. 
     
     
       3. The method of  claim 1 , wherein the portion of the second subset of pharmacy claims are simultaneously provided to the first pharmacy claims approval system. 
     
     
       4. The method of  claim 1 , further comprising:
 routing a portion of the first subset of pharmacy claims to the second pharmacy claims approval system in response to the first set of decisions. 
 
     
     
       5. The method of  claim 1 , wherein the second pharmacy claims approval system includes a machine learning model. 
     
     
       6. The method of  claim 1 , wherein the first set of decisions indicate if the first subset of pharmacy claims is approved. 
     
     
       7. The method of  claim 1 , wherein the second set of decisions indicate if the second subset of pharmacy claims is approved. 
     
     
       8. A system comprising:
 at least one processor; and 
 a memory storing instructions that, when executed by the at least one processor, configure the system to:
 access a plurality of pharmacy claims, wherein each pharmacy claim in the plurality of pharmacy claims includes a set of pharmacy claim data; 
 for each pharmacy claim in the plurality of pharmacy claims, determine a respective aggregate rating of the pharmacy claim based on the set of pharmacy claim data; 
 based on the respective aggregate rating of the plurality of pharmacy claims, selectively process each of the plurality of pharmacy claims, wherein:
 the selectively processing is performed using at least one of:
 a first pharmacy claims approval system including a real-time claim analysis, or 
 a second pharmacy claims approval system including a statistical-model-based claim analysis, and 
 
 the selectively processing includes:
 determining whether a first rating associated with a first subset of pharmacy claims has met a ratings threshold; 
 in response to determining that the first rating has not met the ratings threshold, submitting the first subset of pharmacy claims to the first pharmacy claims approval system; 
 determining whether a second rating associated with a second subset of pharmacy claims has met the ratings threshold; and 
 in response to determining that the second rating associated with the second subset of pharmacy claims has met the ratings threshold, submitting the second subset of pharmacy claims to the second pharmacy claims approval system comprising a neural network; 
 
 
 submit a portion of the second subset of pharmacy claims to the first pharmacy claims approval system; 
 receive from the first pharmacy claims approval system:
 a first set of decisions for the first subset of pharmacy claims, and 
 a set of validation decisions for the portion of the second subset of pharmacy claims; 
 
 store training data including:
 the portion of the second subset of pharmacy claims, and 
 the set of validation decisions; 
 
 train the neural network of the second pharmacy claims approval system based on the training data to generate claims decisions for each of the plurality of pharmacy claims in the training data, wherein:
 the second pharmacy claims approval system is trained to generate a second set of decisions for the second subset of pharmacy claims, and 
 the neural network is trained by performing training operations including:
 accessing the training data; 
 extracting features from the training data to generate predictions by finding correlations among the extracted features of the plurality of pharmacy claims that affect the set of validation decisions, wherein the predictions include a probability that a given pharmacy claim of the plurality of pharmacy claims will be approved; and 
 building the neural network of the second pharmacy claims approval system to generate new predictions for new pharmacy claims in response to analyzing the training data; 
 
 
 receive, from the second pharmacy claims approval system, a second set of decisions for the second subset of pharmacy claims; 
 validate the second set of decisions based on the set of validation decisions; 
 cause display of the first set of decisions and the second set of decisions on a graphical user interface; 
 in response to a determination that a respective claim of the plurality of pharmacy claims has been approved:
 display the respective claim with a first visual indicator, 
 identify a prescription drug associated with the respective claim, and 
 fulfil the prescription drug without operator intervention by:
 automatically controlling movement of a set of containers relative to an automated dispensing device, and 
 automatically dispensing, via the automated dispensing device, the prescription drug into a container of the set of containers; and 
 
 
 in response to a determination that the respective claim has not been approved:
 display the respective claim with a second visual indicator, and 
 display a notification. 
 
 
 
     
     
       9. The system of  claim 8 , wherein the set of pharmacy claim data includes drug data associated with a drug, physician data associated with a physician, pharmacy data associated with a pharmacy, benefit group data associated with a benefit group and member data associated with a member, and any combination thereof. 
     
     
       10. The system of  claim 8 , wherein the portion of the second subset of pharmacy claims are simultaneously provided to the first pharmacy claims approval system. 
     
     
       11. The system of  claim 8 , wherein the instructions configure the system to:
 route a portion of the first subset of pharmacy claims to the second pharmacy claims approval system in response to the first set of decisions. 
 
     
     
       12. The system of  claim 8 , wherein the second pharmacy claims approval system includes a machine learning model. 
     
     
       13. The system of  claim 8 , wherein the first set of decisions indicate if the first subset of pharmacy claims is approved. 
     
     
       14. The system of  claim 8 , wherein the second set of decisions indicate if the second subset of pharmacy claims is approved. 
     
     
       15. A non-transitory computer-readable storage medium, the computer-readable storage medium including instructions that when executed by a computer, cause the computer to:
 access a plurality of pharmacy claims, wherein each pharmacy claim in the plurality of pharmacy claims including a set of pharmacy claim data; 
 for each pharmacy claim in the plurality of pharmacy claims, determine a respective aggregate rating of the pharmacy claim based on the set of pharmacy claim data; 
 based on the respective aggregate rating of the plurality of pharmacy claims, selectively process each of the plurality of pharmacy claims, wherein:
 the selectively processing is performed using at least one of:
 a first pharmacy claims approval system including a real-time claim analysis, or 
 a second pharmacy claims approval system including a statistical-model-based claim analysis, and 
 
 the selectively processing including:
 determining whether a first rating associated with a first subset of pharmacy claims has met a ratings threshold; 
 in response to determining that the first rating has not met the ratings threshold, submit a first subset of pharmacy claims to a first pharmacy claims approval system; 
 determining whether a second rating associated with a second subset of pharmacy claims has met the ratings threshold; and 
 in response to determining that the second rating associated with the second subset of pharmacy claims has met the ratings threshold, submitting the second subset of pharmacy claims to the second pharmacy claims approval system including a neural network; 
 
 
 submit a portion of the second subset of pharmacy claims to the first pharmacy claims approval system; 
 receive, from the first pharmacy claims approval system:
 a first set of decisions for the first subset of pharmacy claims, and 
 a set of validation decisions for the portion of the second subset of pharmacy claims; 
 
 store training data including:
 the portion of the second subset of pharmacy claims, and 
 the set of validation decisions; 
 
 train the neural network of the second pharmacy claims approval system based on the training data to generate claims decisions for each of the plurality of pharmacy claims in the training data, wherein:
 the second pharmacy claims approval system is trained to generate a second set of decisions for the second subset of pharmacy claims, and 
 the neural network is trained by performing training operations including:
 accessing the training data; 
 extracting features from the training data to generate predictions by finding correlations among the extracted features of the plurality of pharmacy claims that affect the set of validation decisions, wherein the predictions include a probability that a given pharmacy claim of the plurality of pharmacy claims will be approved; and 
 building the neural network of the second pharmacy claims approval system to generate new predictions for new pharmacy claims in response to analyzing the training data; 
 
 
 receive, from the second pharmacy claims approval system, a second set of decisions for the second subset of pharmacy claims; 
 validate the second set of decisions based on the set of validation decisions; 
 cause display of the first set of decisions and the second set of decisions on a graphical user interface; 
 in response to determining that a respective claim of the plurality of pharmacy claims has been approved:
 display the respective claim with a first visual indicator, 
 identify a prescription drug associated with the respective claim, and 
 fulfill the prescription drug without operator intervention by:
 automatically controlling movement of a set of containers relative to an automated dispensing device, and 
 automatically dispensing, via the automated dispensing device, the prescription drug into a container of the set of containers; and 
 
 
 in response to determining that the respective claim has not been approved:
 display the respective claim with a second visual indicator, and 
 display a notification. 
 
 
     
     
       16. The non-transitory computer-readable storage medium of  claim 15 , wherein the set of pharmacy claim data includes drug data associated with a drug, physician data associated with a physician, pharmacy data associated with a pharmacy, benefit group data associated with a benefit group and member data associated with a member, and any combination thereof. 
     
     
       17. The non-transitory computer-readable storage medium of  claim 15 , wherein the portion of the second subset of pharmacy claims are simultaneously provided to the first pharmacy claims approval system.